Recent research has documented a higher miscarriage risk among sexual minority individuals than heterosexual individuals. These disparities may be partly mediated by age at pregnancy, gravidity, and use of medically assisted reproduction (MAR). We used data from four cohorts (Nurses' Health Study 2 and 3, Growing Up Today Study, and National Longitudinal Study of Adolescent to Adult Health) to estimate randomized interventional analogues of mediational direct and indirect effects of sexual orientation on miscarriage (pure natural direct effects [rPNDE], total natural indirect effects [rTNIE], and proportions mediated [rPM]). We found that pregnancies to sexual minority individuals were more likely to end in a miscarriage than those to heterosexual individuals. We found weak mediation of this effect by age (rPNDE: 1.23, 95% confidence interval [CI]: 1.15-1.32; rTNIE: 1.02, 95% CI: 1.01-1.03; rPM: 12%, 95% CI: 5%-19%), but not gravidity (rPM: 2%, 95% CI: -1%-6%) and MAR use (rPM: -1%, 95% CI: -4%-3%). While these factors impact miscarriage risk, our findings suggest that sexual orientation disparities in miscarriage are only minimally explained by age, and not explained by gravidity and use of MAR, contributing to the understanding of the role of pregnancy-related factors vs. structural factors in sexual orientation disparities in miscarriage.